name: conversion-and-io
description: "Convert Earth Engine JavaScript, Python, notebooks, exports, and
geospatial data formats with geemap."
disable-model-invocation: true
metadata:
disco-role: operating
license: MIT
conversion-and-io
Use this sub-skill when the task is about moving code or data with geemap:
Earth Engine JavaScript/Python/notebook conversion, local vector/raster format
conversion, Earth Engine image/vector/video exports, NumPy/xarray extraction,
COG/STAC/titiler tile URLs, OSM wrappers, or data-catalog search.
First choices
- For JavaScript/Python/notebook conversion, read
references/workflows.md
and use the bundled scripts/convert_ee_js.py
when a file/folder conversion is requested.
- For export planning, read
references/workflows.md
and use scripts/export_task_checklist.py
before launching Drive/Asset/Cloud Storage tasks or large local downloads.
- For file formats, CRS requirements, selectors, and optional dependencies, read
references/data-formats.md.
- For API names and signatures, read
references/api-reference.md.
- For Earth Engine authentication, project setup, and package installation,
use the root installation-and-auth reference.
- For failures, missing extras, or stalled tasks, read
references/troubleshooting.md.
Boundary and routing
Stay here for:
geemap.conversion: js_to_python, js_snippet_to_py,
js_to_python_dir, py_to_ipynb, py_to_ipynb_dir,
execute_notebook, get_js_examples, get_nb_template.
geemap.common / geemap.coreutils: ee_export_image,
ee_export_image_collection, ee_export_vector, export-to-Drive/Asset/Cloud
Storage helpers, ee_to_numpy, ee_to_xarray, csv_to_geojson,
csv_to_shp, shp_to_ee, ee_to_geojson, zonal_stats, cog_tile,
stac_tile, and coreutils.geojson_to_ee.
- Optional OSM downloads through
geemap.osm when osmnx/geopandas and
network access are available.
- Data catalog search helpers such as
geemap.search_ee_data and
geemap.datasets.DATA when the user needs Earth Engine asset IDs.
Route elsewhere for:
Safety model
- Local conversion rewrites files or notebooks and usually does not contact
Earth Engine unless notebooks are executed.
- Local EE downloads (
ee_export_image, ee_export_vector, zonal_stats,
ee_to_numpy) contact Earth Engine immediately and require initialization,
network access, and a bounded region/scale.
- Batch exports (
*_to_drive, *_to_asset, *_to_cloud_storage) start
asynchronous Earth Engine tasks; task completion is delayed and must be
monitored outside Python.
- COG/STAC/OSM/data catalog helpers usually contact third-party HTTP
services and may need proxy handling.
- Optional extras (
geopandas, osmnx, geedim, localtileserver, xee,
gdown, ffmpeg, ipynb-py-convert) should be called out before use.